传统彩色图像二值化是先将彩色图像转换成黑白图像后，再对黑白图像进行的二值化。在转换过程中有

可能丢失原始图像信息。提出的基于图像特征值的彩色图像二值化是基于原始图像特征值进行的二值化，提出了先粗分后细分的逻辑规范。粗分时，先根据特征表达式，计算图像的特征值阈值，然后以阈值对图像像素进行划分；细分时，依据像素颜色相近，对图像像素进一步细分，能让图像二值化达到比较好的效果

-After conventional color image binarization is the first color images into black and white images, and then the black and white image binarization. The original image information may be lost in the conversion process. Proposed a color image based on binary image feature values is carried out based on the original image feature value binarization, proposed after the first rough segmentation logic specification segments. Crude sharing, according to the first characteristic expression, calculates the image characteristic value of the threshold, then the threshold of the image pixel is divided　when broken down, according to the similar color of pixels on the image pixel is further broken down, allowing binary image to achieve more good results